Towards VO 2 monitoring: Validation of a heart rate based algorithm

This article presents and validates a novel algorithm for the continuous monitoring of the VO 2 during exercise. The algorithm relies on instantaneous HR measurements to provide a continuous estimation, and can be integrated in a wearable device (e.g., smartwatch, sensor patch). It can be customized by user's main anthropomorphic parameters and automatically learns from newly incoming data recalibrating itself if needed. The system is evaluated against a database of 14 healthy subjects performing various maximal endurance tests. The proposed method provides a VO 2 estimation with average RMSE of 4.63 ml/kg/min.